The long-term goal of the present project is to understand the neuronal activity underlying the process of voluntary control. Historically, investigating this process has been constrained largely by the fact that voluntary motor output is naturally coupled to motion of a body part, to the muscle contractions moving that body part, and to the sensory feedback produced by the motion of that body part. Now, as knowledge of these relationships is being harnessed to control brain computer interfaces (BCIs), BCIs themselves provide a new paradigm for directly examining the neuronal processes underlying voluntary control. As the brain controls a BCI, neuronal activity becomes dissociated from movement of the body and devoted instead to voluntary control of the interface. Movement of the native limb may cease, and EMG activity may be absent as neurons continue to control the BCI voluntarily. Hence proprioceptive feedback and visual observation of limb movement may be absent as well. Carefully chosen BCI paradigms thus provide an unprecedented opportunity to examine voluntary control of neuronal activity per se, dissociated from motor output and sensory feedback. Here we propose to investigate the neuronal processes underlying voluntary control using a simple BCI paradigm that assesses the single-session performance of neurons in voluntarily controlling a novel interface. Our BCI paradigm assesses the ability to coordinate the activity of small ensembles of arbitrarily-selected neurons in novel patterns. Specifically, th present proposal aims to determine whether the brain's ability to control neurons voluntarily depends: i) on the cortical area (motor, premotor, and parietal areas will be compared), ii) on the presence or absence of visual and/or somatosensory inputs, and iii) on output projections to different levels of the neuraxis (neurons with cortico-cortical axons, axons projecting to the brainstem, cortico-spinal axons, and cortico-motoneuronal connections will be compared). Current efforts at neuro-prosthetic control of artificial hands, while impressive, have not progressed as rapidly as might have been expected. In part this may reflect inadequate basic understanding of the neuronal activity underlying the process of voluntary control per se. Thus, improved understanding of this fundamental process will lead both to improved neuro-prosthetic devices for restoration of lost function and to improved neuro-rehabilitation for functional recovery in patients affected by a wide variety of neurological diseases including stroke, amyotrophic lateral sclerosis, multiple sclerosis, brain or spinal cord injury, and cerebral palsy.